We thought automation was the hard part. It wasn’t.

We thought automation was the hard part. It wasn’t.

Getting a workflow to run once is easy.

The real challenge starts when people need it to run:

• repeatedly

• across devices

• across different states

• without guessing what broke

For a while, we kept thinking users wanted more features.

More automation.
More AI.
More ways to trigger things.

But the more we built, the clearer it became:

what people actually needed was more visibility and control.

They wanted to know:

• what state the device was in

• whether the task was still running

• whether something failed silently

• what to restart

• what was being occupied

• what was actually reliable at scale

That changed how we started building.

Less “look what it can do.”
More “can people trust it when they use it every day?”

Curious if other builders here have gone through the same shift:

thinking the problem was one thing, then realizing the real bottleneck was something much less glamorous?